Selective Heteroaryl N-Oxidation of Amine-Containing Molecules

Article abstract of DOI:10.1021/acs.orglett.8b00558

The first examples of nonenzymatic N-oxidation of heteroarenes in the presence of amines are reported. Pyridine, quinoline, and isoquinoline N-oxides are selectively formed in the presence of more reactive aliphatic and alicyclic amines by use of an in situ protonation strategy and an iminium salt organocatalyst. Application to late-stage functionalization that mimics phase 1 metabolism of small-molecule drugs is also demonstrated.

Full text of DOI:10.1021/acs.orglett.8b00558

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Selective Heteroaryl N‑Oxidation of Amine-Containing Molecules
Robert M. B. Dyer, Philip L. Hahn, and Michael K. Hilinski*
Department of Chemistry, University of Virginia, Charlottesville, Virginia 22904-4319, United States
S
* Supporting Information
ABSTRACT: The first examples of nonenzymatic N-
oxidation of heteroarenes in the presence of amines are
reported. Pyridine, quinoline, and isoquinoline N-oxides are
selectively formed in the presence of more reactive aliphatic
and alicyclic amines by use of an in situ protonation strategy
and an iminium salt organocatalyst. Application to late-stage
functionalization that mimics phase 1 metabolism of small-molecule drugs is also demonstrated.
Scheme 1. Site-Selective Oxidation of N-Heteroarenes
chieving enzymatic levels of selectivity in oxidation
Areactions without resorting to biocatalysis is a significant
unsolved problem in synthetic methods development.1,2 In
particular, the selective late-stage oxidation of natural products
and designed bioactive molecules in order to modulate
biological activity and physicochemical properties is regarded
as an enabling application.3 One of the biggest unsolved
challenges in this area is the development of nonenzymatic
methods that directly mimic oxidative metabolism.4 This is of
critical importance to the pharmaceutical industry, given the
need for rapid identification, preparation, and characterization
of the biological activity of drug metabolites.5 Due to the
relative structural complexity of drug candidates, which
frequently include multiple functional groups that are prone
to oxidation, the desired selectivity can be difficult to attain.
One class of transformations for which this is the case are
metabolic oxidations of nitrogen-containing heteroaromatic
rings (Scheme 1a). These generally fall into two categories: N-
oxidation catalyzed most often by the cytochrome p450
(CYP)6,7 or flavin-containing monooxygenase (FMO)8 families
of enzymes and C(sp2)−H hydroxylation adjacent to nitrogen
catalyzed most often by aldehyde oxidase (AO).9 In regards to
the former transformation, nonenzymatic site-selective hetero-
aryl N-oxidation is complicated by the fact that pharmaceuticals
often contain aliphatic amines, which are generally more
nucleophilic and thus are preferred sites of oxidation. At
present, this complication can only be dealt with indirectly. For
example, the state of the art for achieving selective synthesis of
heteroaryl N-oxides in tertiary amine-containing compounds is
exhaustive N-oxidation followed by selective reduction
(Scheme 1b).10,11 This approach, first illustrated in the 1940s,
has not been demonstrated to be widely applicable. Given that
more than two potential N-oxidation sites in a substrate or
functionality sensitive to reducing conditions would greatly
complicate this process, a one-step site-selective solution would
be the preferred contemporary strategy. However, no examples
of such a transformation have been reported. In addition to
metabolite synthesis, methods for predictably site-selective N-
oxidation would have other potential applications in drug
discovery, given the increasing attention paid in recent years to
heteroaromatic N-oxides as potential drug candidates.12 As
such, the lack of any nonenzymatic site-selective solutions
creates a barrier to the development of new therapeutics.
Herein, we describe unprecedented site-selective N-oxidations
of heteroarenes in the presence of aliphatic amines. We also
demonstrate the application of this method to late-stage
oxidation that mimics xenobiotic metabolism.
Our strategy for site-selective N-oxidation takes advantage of
the fact that, in general, the reactivity of amines and
heteroarenes toward electrophilic oxidants has a positive
Received: February 14, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b00558
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
correlation with basicity. In the presence of a suitable oxidant,
in situ protection of a more nucleophilic amine by use of a
Brønsted acid would in principle lead to selective formation of
the heteroaromatic N-oxide.13 For typical drug substrates
containing an amine and a heteroarene, such as nicotine (1,
Table 1), the difference in acidity between the protonated
hydrogen peroxide employed (2 equiv). No pyrrolidine N-
oxide was observed under these conditions, highlighting the
considerable site selectivity enabled by the protonation strategy.
Encouraged by these results, we evaluated whether the prior
preparation of the amine salt could be avoided, enabling the
desired one-step site-selective transformation. Addition of 1
equiv of HBF4 to the iminium salt-catalyzed oxidation
conditions enabled selective oxidation of the nicotine free
base, with isolated yield identical to that achieved using the pre-
prepared salt (entry 5). Control reactions indicate that catalyst
4 is required (entry 6) and that the use of HFIP, which we have
previously shown is required as a cosolvent for best results
when using catalyst 417 and which is separately known to
modulate oxidation selectivity through hydrogen bonding
effects,18 is not sufficient to provide the desired selectivity in
the absence of an appropriate oxidant (entry 7) or HBF4 (entry
8). Strong acids other than HBF4 (e.g., H2SO4) can be used
with comparable results (entry 9). Finally, under the optimized
conditions the use of an organocatalytic strategy offers superior
results to the use of isolated DMDO (entry 10).
Pyridines are the most abundant nitrogen-containing
aromatic heterocycles in approved drugs19 and are frequent
targets for metabolic N-oxidation.6,8 We evaluated the substrate
scope of site-selective N-oxidation by employing a number of
differentially substituted pyridine-containing substrates, which
in each case also contained at least one aliphatic amine (Table
2). In all but one case, site-selective pyridine N-oxidation was
observed, and in many cases, synthetically useful yields were
achieved. In all cases, yields were limited primarily by
conversion rather than substantial formation of the undesired
N-oxide. Alicyclic, tertiary, and secondary amines were
tolerated; however, attempts to oxidize a substrate bearing a
primary amine (11) resulted in trace amounts of oxidation of
the primary amine and no conversion to the desired pyridine
N-oxide. Substitution of the pyridine ring at the 2-position
proved detrimental compared to substitution at the 3- or 4-
position (substrate 27 vs 25 or 5), which could arise from steric
effects or decreased nucleophilicity of the pyridine nitrogen due
to inductive withdrawing effects of the protonated amine.
Oxidation of a substrate containing a piperazine ring (19) gave
selectively the product of pyridine oxidation, demonstrating
that site selectivity is also achievable in the presence of more
than one tertiary amine nitrogen (entry 6). In this case 1 equiv
of HBF4 was sufficient to prevent more than trace oxidation of
either piperazine nitrogen.
a
Table 1. Optimization of the Reaction Conditions
ratio of
b
yield of
c
a
entry substrate
conditions
1:2:3
3 (%)
1
2
3
4
5
1·HBF4
1·HBF4
1·HBF4
1·HBF4
1
m-CPBA, CH2Cl2, rt
H2O2, AcOH, 80 °C
DMDO, CH2Cl2, rt
H2O2, 4 (20 mol %)
1:5:0
0:1:0
11:0:1
1:0:5
1:0:4



77%
77%
HBF4·OEt2, H2O2,
4 (20 mol %)
6
7
8
9
1
1
1
1
HBF4·OEt2, H2O2
HBF4·OEt2, m-CPBA
H2O2, 4 (20 mol %)
1:0:0
2:5:0
0:1:0




H2SO4, H2O2,
4 (20 mol %)
66%
10
1
HBF4·OEt2, DMDO

15%
a
One equivalent of oxidant was used for entries 1, 3, 7, and 10. Two
equivalents were used for entries 4−6, 8, and 9. Four equivalents were
used for entry 2. For entries 4−10, a 5:1 mixture of CH2Cl2:HFIP was
used as the solvent, and the reactions were run at room temperature.
b
c
Ratios determined by integration of HPLC chromatograms. Isolated
yield.
amine and protonated heteroarene is on the order of 5 pKa
units, which would correspond to a 105-fold difference in
concentration between the two protonated forms when 1 equiv
of a strong acid is used. Thus, identification of oxidation
conditions compatible with use of a stoichiometric acid additive
and that demonstrate a sufficient rate of heteroarene oxidation
would in principle lead to the desired selectivity.
Experiments were also performed to probe the effect of
additional substitution on the pyridine ring, which revealed
additional functional group compatibility and the influence of
substituent effects on reaction outcome (Table 3). Pyridines
substituted with halogens in the 3-position gave the highest
yields observed for any substrates (entry 1). In contrast,
strongly σ-withdrawing substituents at the 2-position (F, CF3)
resulted in no observed product formation.20 Other 2-
substituted pyridines (OMe, CH3) were oxidized in good
yields (entry 2). Site-selective N-oxidation was also observed
for substrates bearing quinoline and isoquinoline rings (entries
3−6), expanding the range of pharmaceutically relevant
heteroaromatic scaffolds shown to be amenable to this
chemistry.
Initial evaluation of this strategy using the isolated mono-
HBF4 salt of nicotine as a substrate revealed that peracids,
historically oxidants of choice for pyridine N-oxidation,11,14
provided only the undesired pyrrolidine N-oxide when
employed under typical conditions (Table 1, entries 1 and 2).
Alternatively, the use of dimethyldioxirane (DMDO)15
provided the first indication of the desired selectivity, resulting
in exclusive formation of pyridine N-oxide 3, albeit at low
conversion (11:1 ratio of starting material to desired product)
when 1 equiv was employed (entry 3). To avoid using a large
excess of reagent, we explored the use of organocatalytic
conditions known to approximate the reactivity of DMDO.16
Specifically, using a trifluoromethyl iminium salt developed
previously in our laboratory for use in catalytic oxaziridinium-
mediated oxidations17 the desired conversion and selectivity
were attained, providing 77% isolated yield of free base 3
following basic workup and purification (entry 4). The use of
10−20 mol % of 4 allowed for minimization of the amount of
For complex, drug-like, or natural product substrates, the
products of site-selective N-oxidation can in principle be
subjected to further known transformations of N-oxides,21
enhancing the impact of this single transformation by providing
B
DOI: 10.1021/acs.orglett.8b00558
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
a
a
Table 2. Scope of the Amine Component
Table 3. Scope of the Heteroarene Component
a
Reaction conditions: substrate (0.5 mmol), 4, (0.1 mmol), H2SO4
a
Reaction conditions: substrate (0.5 mmol), 4, (0.05 mmol), HBF4·
OEt2 (0.5 mmol), and 50% aq. H2O2 (1.0 mmol) in a mixture of
CH2Cl2 (2.5 mL) and HFIP (0.5 mL) under air atmosphere. Entry 6:
H2SO4 (0.5 mmol) was used in place of HBF4·OEt2. Entry 4: Reaction
(0.5 mmol), and 50% aq. H2O2 (1.0 mmol) in a mixture of CH2Cl2
b
(2.5 mL) and HFIP (0.5 mL) under air atmosphere. Isolated yield.
c
HBF4·OEt2 (0.5 mmol) was used in place of H2SO4.
b
performed on 1.5 mmol scale. Isolated yield.
access to additional diverse products of late-stage functionaliza-
tion. As one application of this strategy, we envisioned that
formal site-selective C(sp2)−H hydroxylation might be
achieved by a two-step sequence of selective N-oxidation
followed by subsequent transformation to the desired formal
C−H hydroxylation product. We specifically selected this
transformation for its potential ability to mimic metabolic
oxidation of bioactive compounds by aldehyde oxidase (AO),22
for which there is no analogous nonenzymatic method. The
importance of AO metabolism and its impact on drug discovery
have become increasingly recognized in recent years,9 with
several clinical drug failures resulting from metabolism by AO
in humans that had not been observed in preclinical animal
studies.23 Quinine (51), a known substrate of AO,24 was
subjected to this two-step C−H hydroxylation strategy. First,
the desired quinoline N-oxide was formed selectively in 54%
yield (Scheme 2). This reaction highlights the functional group
tolerance of the catalytic N-oxidation, given that the desired
product is selectively formed in the presence of other easily
oxidized functional groups, such as an olefin and a secondary
benzylic alcohol, in addition to a tertiary amine. The N-oxide
C
DOI: 10.1021/acs.orglett.8b00558
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 2. Two-Step Strategy to Mimic Metabolic C(sp3)−H
Hydroxylation by Aldehyde Oxidase
(6) Lewis, D. F. V. Curr. Med. Chem. 2003, 10, 1955−1972.
(7) Vitaku, E.; Smith, D. T.; Njardarson, J. T. J. Med. Chem. 2014, 57,
10257−10274.
(8) Krueger, S. K.; Williams, D. E. Pharmacol. Ther. 2005, 106, 357−
387.
(9) Pryde, D. C.; Dalvie, D.; Hu, Q.; Jones, P.; Obach, R. S.; Tran, T.-
D. J. Med. Chem. 2010, 53, 8441−8460.
(10) Ochiai, E.; Okamoto, T.; Kobayashi, Y. Yakugaku Zasshi 1948,
68, 109−111.
(11) Taylor, E.; Boyer, N. J. Org. Chem. 1959, 24, 275−277.
(12) Mfuh, A. M.; Larionov, O. V. Curr. Med. Chem. 2015, 22, 2819−
2857.
(13) For selected examples of this strategy as applied to other
selective oxidations of amine substrates see: (a) Asensio, G.; Gonzalez-
Nunez, M. E.; Bernardini, C. B.; Mello, R.; Adam, W. J. Am. Chem. Soc.
52 can be converted in one step to the observed product of AO
oxidation as shown by Cook and co-workers.25 Overall, this
oxidation/transposition sequence demonstrates both the
selectivity and valuable application of this method to metabolite
synthesis.
́
1993, 115, 7250−7253. (b) Ferrer, M.; Sanchez-Baeza, F.; Messeguer,
A.; Diez, A.; Rubiralta, M. J. Chem. Soc., Chem. Commun. 1995, 50,
293−294. (c) Lee, M.; Sanford, M. S. J. Am. Chem. Soc. 2015, 137,
12796−12799. (d) Howell, J. M.; Feng, K.; Clark, J. R.; Trzepkowski,
L. J.; White, M. C. J. Am. Chem. Soc. 2015, 137, 14590−14593.
(e) Mbofana, C. T.; Chong, R.; Lawniczak, J.; Sanford, M. S. Org. Lett.
2016, 18, 4258−4261. (f) Lee, M.; Sanford, M. S. Org. Lett. 2017, 19,
572−575. (g) Mack, J. B. C.; Gipson, J. D.; Du Bois, J.; Sigman, M. S.
J. Am. Chem. Soc. 2017, 139, 9503−9506.
In summary, we report the first demonstration of a method
for heteroarene N-oxidation that is both predictably site
selective and suitable for late-stage functionalization. The
method reverses the inherent N-oxidation selectivity for all
substrates evaluated, favoring heteroarene oxidation over the
oxidation of aliphatic amines. As such, it is expected to provide
a platform for selective modification of complex bioactive
compounds that is complementary to other approaches.
Furthermore, the demonstrated application of this method to
metabolite synthesis could facilitate drug discovery in cases
where the rapid preparation of relevant metabolites is
unattainable through other means.
(14) Cymerman, C. J.; Purushothaman, K. K. J. Org. Chem. 1970, 35,
1721−1722.
(15) Curci, R.; Dinoi, A.; Rubino, M. F. Pure Appl. Chem. 1995, 67,
811−822.
(16) Typical dioxirane-mediated catalytic conditions require basic pH
which would be incompatible with the protonation strategy. See the
following and references therein: Shuler, W. G.; Johnson, S. L.;
Hilinski, M. K. Org. Lett. 2017, 19, 4790−4793.
ASSOCIATED CONTENT
* Supporting Information
■
S
(17) Wang, D.; Shuler, W. G.; Pierce, C. J.; Hilinski, M. K. Org. Lett.
2016, 18, 3826−3829.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.8b00558.
́
(18) Dantignana, V.; Milan, M.; Cusso, O.; Company, A.; Bietti, M.;
Costas, M. ACS Cent. Sci. 2017, 3, 1350−1358.
(19) Vitaku, E.; Smith, D. T.; Njardarson, J. T. J. Med. Chem. 2014,
57, 10257−10274.
Experimental details as well as spectroscopic and analytic
data for all new compounds and oxidation products
(PDF)
(20) Evidence suggests that pyridine N-oxides bearing strongly
electron-withdrawing groups at the 2-position are unstable. See:
Sarantakis, D.; Sutherland, J. K.; Tortorella, C.; Tortorella, V. J. Chem.
Soc. C 1968, 72−73.
AUTHOR INFORMATION
■
(21) Bull, J. A.; Mousseau, J. J.; Pelletier, G.; Charette, A. B. Chem.
Rev. 2012, 112, 2642−2713.
Corresponding Author
*E-mail: hilinski@virginia.edu
ORCID
(22) Garattini, E.; Terao, M. Expert Opin. Drug Metab. Toxicol. 2012,
8, 487−503.
(23) Zhang, X.; Liu, H.-H.; Weller, P.; Zheng, M.; Tao, W.; Wang, J.;
Liao, G.; Monshouwer, M.; Peltz, G. Pharmacogenomics J. 2011, 11,
15−24.
Michael K. Hilinski: 0000-0003-2861-7099
Notes
(24) Lepri, S.; Ceccarelli, M.; Milani, N.; Tortorella, S.; Cucco, A.;
Valeri, A.; Goracci, L.; Brink, A.; Cruciani, G. Proc. Natl. Acad. Sci. U. S.
A. 2017, 114, E3178−E3187.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
(25) Díaz-Arauzo, H.; Cook, J. M.; Christie, D. J. J. Nat. Prod. 1990,
́
Funding from the American Chemical Society Petroleum
Research Fund Doctoral New Investigator Program
(PRF#56158-DNI), the National Institutes of Health (R01
GM124092), and the University of Virginia is gratefully
acknowledged.
53, 112−124.
REFERENCES
■
(1) Newhouse, T.; Baran, P. S. Angew. Chem., Int. Ed. 2011, 50,
3362−3374.
(2) White, M. C. Science 2012, 335, 807−809.
(3) Michaudel, Q.; Journot, G.; Regueiro-Ren, A.; Goswami, A.; Guo,
Z.; Tully, T. P.; Zou, L.; Ramabhadran, R. O.; Houk, K. N.; Baran, P.
S. Angew. Chem., Int. Ed. 2014, 53 (45), 12091−12096.
́
(4) Genovino, J.; Sames, D.; Hamann, L. G.; Toure, B. B. Angew.
Chem., Int. Ed. 2016, 55, 14218−14238.
(5) Obach, R. S. Pharmacol. Rev. 2013, 65, 578−640.
D
DOI: 10.1021/acs.orglett.8b00558
Org. Lett. XXXX, XXX, XXX−XXX